Biomedical engineers at RMIT University have unveiled a groundbreaking smartphone tool equipped with AI capabilities that could revolutionize the early detection of strokes. This innovative technology uses AI to analyze facial expressions, assisting paramedics in quickly identifying stroke symptoms, which could save lives and prevent long-term disability.

Strokes occur when blood flow to the brain is obstructed, depriving brain cells of oxygen and nutrients. As a leading cause of disability and death worldwide, timely intervention during a stroke is critical. Every minute counts and can significantly influence a patient’s recovery.

“Early detection of stroke is critical, as prompt treatment can significantly enhance recovery outcomes, reduce the risk of long-term disability, and save lives,” emphasized Professor Dinesh Kumar from RMIT’s School of Engineering, who led the research. Under Professor Kumar’s guidance, PhD scholar Guilherme Camargo de Oliveira spearheaded the development of this revolutionary technology. The software employs AI algorithms to scrutinize facial symmetry and muscle movements associated with strokes.

Although not intended to replace comprehensive clinical assessments, the smartphone tool boasts an impressive 82% accuracy rate in identifying stroke symptoms. “Our face-screening tool complements existing diagnostic methods by offering a rapid initial assessment,” explained Kumar.

The technology operates by employing facial expression recognition, a method that evaluates asymmetry and changes in facial muscle actions known as action units. Leveraging the Facial Action Coding System (FACS), originally devised in the 1970s, the tool meticulously categorizes and analyzes facial movements indicative of stroke. “One of the key parameters that affect people with stroke is that their facial muscles typically become unilateral, so one side of the face behaves differently from the other side of the face,” de Oliveira explained.

The study involved video recordings of facial expression examinations from individuals post-stroke and healthy controls, which were crucial for refining and validating the tool’s effectiveness. Recognizing stroke symptoms can be challenging, particularly in diverse patient populations where symptoms may present differently or be overlooked altogether.

“Studies indicate that nearly 13% of strokes are missed in emergency departments and at community hospitals, while 65% of patients without a documented neurological examination experience undiagnosed stroke,” Kumar pointed out. Moreover, subtle signs of stroke can be missed, especially when attending to patients of different racial or gender backgrounds, Kumar added.

The team aims to transform the prototype into a user-friendly smartphone application. They are working with healthcare providers to integrate the tool into current emergency response procedures, helping first responders more effectively identify strokes and other neurological conditions that affect facial expressions.

“We want to be as sensitive and specific as possible. We are now working towards an AI tool with additional data and where we are going to be considering other diseases as well,” remarked Kumar on future plans.

By Impact Lab